We are investigating the in vitro splicing of a class II yeast mitochondrial intron, oxi3 intron-5-gamma (I5g). Precursors containing 15g undergo a self-splicing reaction to yield an intron lariat RNA plus the spliced exons. One aspect of this project involves study of the products of the self-splicing reaction in detail. We propose to map the in vitro and in vivo splice junctions. The composition and location of the branched oligonucleotide component of the lariat will be determined by a combination of nuclease digestion, primer extension, and sequencing methods. Additional experiments are proposed to characterize the products of the I5g reaction under high salt conditions. These products include two that have been tentatively identified as excised 5' and 3' exons which are not ligated. Another may be a 2/3's molecule which could represent an intermediate of the self-splicing reaction. We will continue to investigate the properties of I5g splicing by perturbing the reaction conditions and characterizing altered products. Another part of this project will map regions of I5g which are important for carrying out the rearrangement reaction. These sequences will be identified by examining the ability of derivatives to undergo splicing or the partial reaction of exon cleavage. Truncated versions of the precursor can be made simply by using a shortened template for transcription. Other derivatives will be made with short sequences of DNA inserted at the available restriction sites, and deletions will be constructed in this survey of required sequences. The role of exon sequences will be examined by constructing versions of the precursor with plasmid sequences in place of either 5' or 3' exons. Sequences required in vivo for splicing will be identified by subcloning and sequencing the existing cis-acting mutants of this intron. These will also be tested for in vitro activity. The self-splicing ability of other class II introns from yeast mitochondria and other biological sources will be examined. This will test the idea that self-splicing is a general feature of class II introns. These results also will put us in a position to examine the role of accessory proteins ("maturases") in the splicing of class II introns in vivo and in vitro.